Drimva

Thinking Machines Lab

Model Designer

$300K–$400K

San Francisco · On-site · Full-time · Posted Sep 30, 2026

Experience
Several years in a role focused on how people interact with technology.
Visa sponsorship
Visa sponsorship is offered; success is not guaranteed for every candidate or role.

About the job

Shapes AI model behavior and human-AI interactions, working with researchers and cross-functional teams to improve models, interfaces, data, and evaluations.

What you'll do

  • Translate design judgment into training data, evaluations, and changes to model behavior; design how models collaborate with people.
  • Partner across research, product, design, and engineering to develop interfaces, including live and multimodal interaction.
  • Collect high-quality data, define evaluations of human-model outcomes, and identify improvements from user feedback, quantitative insight, and research priorities.

What they ask for

  • Clear communication, including explaining complex technical concepts in writing.
  • Ability to understand users with needs and contexts different from one’s own, developed through work across varied populations, domains, or cultures.
  • Several years of experience in a role focused on how people interact with technology, such as product design, UX research, content or conversation design, or model behavior work at an AI lab.

Nice to have

  • Experience designing model behavior or personality, shaping model approaches to research and judgment, or influencing technical roadmaps in judgment-dependent domains.
  • Hands-on experience with large language models, including prompting and evaluating outputs; experimentation, quantitative analysis, or a portfolio demonstrating craft and clarity are also welcomed.

Pay

$300,000–$400,000 USD (base salary)

Tools

large language models

AI focus

Shapes model behavior, collaboration, judgment, uncertainty, values, training data, evaluations, and human-AI interfaces, including live and multimodal interaction.

Model-written summary · the original posting is authoritative

First seen Oct 10, 2026 · Last verified Nov 5, 2026 (UTC)